LAUSR

Based on the results of this multicenter study of severe liver injuries from 10 trauma centers across North America, we recommend that ERCP be reserved for bile leaks with >300 ml of bilious drain output. Smaller leaks can be managed with percutaneous drainage alone. BACKGROUND Optimal management of bile leaks (BLs) after severe liver injury is unknown. Study objectives were to define current practices in diagnosis and management of BL to determine which patients may benefit from endoscopic retrograde cholangiopancreatography (ERCP). METHODS American Association for the Surgery of Trauma grade ≥III liver injuries from 10 North American trauma centers were included in this retrospective study (February 2011 to January 2021). Groups were defined as patients who developed BL versus those who did not. Subgroup analysis of BL patients was performed by management strategy. Bivariate analysis compared demographics, clinical/injury data, and outcomes. Receiver operating characteristic curves were performed to investigate the relationship between bilious drain output and ERCP. RESULTS A total of 2,225 patients with severe liver injury met the study criteria, with 108 BLs (5%). Bile leak patients had higher American Association for the Surgery of Trauma grade of liver injury (p < 0.001) and were more likely to have been managed operatively from the outset (69% vs. 25%, p < 0.001). Bile leak was typically diagnosed on hospital day 6 [4–10] via surgical drain output (n = 37 [39%]) and computed tomography scan (n = 34 [36%]). On the BL diagnosis day, drain output was 270 [125–555] mL. Endoscopic retrograde cholangiopancreatography was the most frequent management strategy (n = 59 [55%]), although 32 patients (30%) were managed with external drains alone. Bile leak patients who underwent ERCP, surgery, or percutaneous transhepatic biliary drain had higher drain output than BL patients who were managed with external drains alone (320 [180–720] vs. 138 [85–330] mL, p = 0.010). Receiver operating characteristic curve analysis of BL demonstrated moderate accuracy (area under the receiver operating characteristic curve, 0.636) for ERCP at a cutoff point of 390 mL of bilious output on the day of diagnosis. CONCLUSION Patients with BL >300 to 400 mL were most likely to undergo ERCP, percutaneous transhepatic biliary drain, or surgical management. Once external drainage of BL has been established, we recommend ERCP be reserved for patients with BL >300 mL of daily output. Prospective multicenter examination will be required to validate these retrospective data. LEVEL OF EVIDENCE Therapeutic and Care Management; Level IV.